株式会社primeNumber 社名: 2015年11月 設立: - to B のデータ領域のサービスを展開 - 2018年末「trocco®」リリース - 2019年7月シリーズAラウンド資金調達 Data Engineering Study #1 2020/07/15(水) 19:30〜 ForkewllさんとprimeNumberの共催
Output File base and developed by Treasure Data(sakama .et al) on Mar 2015 v0.3.0.pre1 is JRuby Output base and developed by Sonots on Mar 2016 embulk-output-bigquery_java is Java Output base, is compatible with JRuby version base, and developed by trocco from Feb 2020 Our motivation 1. hit the limit of transfer time for customer requirement using JRuby 2. minimizing transfer time reduces our infrastructure cost. most of customers use BigQuery as DWH
problem(https://twitter.com/hiroysato/status/1269822234454982656) google-api-client incompatibility JRuby and JRE related problem (https://github.com/embulk/embulk-output-bigquery/issues/122) Sometime error: output plugin bigquery not found embulk/embulk#1214 Sometimes failed to run(can not copy embedded jar to temp directory) embulk/embulk#1148 Occasionally failed by "JRuby runtime is not loaded successfully" error embulk/embulk#978
replace embulk/embulk-output-bigquery and release 0.7.0 in embulk/embulk-output-bigquery at Spring 2021 drop some options related JRuby performance issues: e.g. payload_column_index number of contributor: 1 only me, the more contributors joins this project, the faster we can achieve milestone in road map :)